George Mills wrote: > > Here is the "on demand" cache version I spoke of. > > Note it runs slower (because it should). > > Again it depends on your applications bevahvior as to its worth. > It would payoff if the factorials needed were sparse (but still > repeatedly used [multiple hits in the cache]). > > If I made FACT any slower Erich's post might start to payoff :-) > > to do > make "start timemilli > invalidate_fact_cache > for[num 100 999][ > if equalp :num(sum fact item 1 :num > fact item 2 :num > fact item 3 :num) > [pr :num]] > show (timemilli - :start) / 600 > end > > to fact :n > if (item :n :factarr) = -1 [setitem :n :factarr fact_real :n] > op item :n :factarr > end > > to fact_real :n > op ifelse :n<2[1][:n*fact_real :n-1] > end > > to invalidate_fact_cache > make "factarr(array 10 0) > for[i 0 9][ > setitem :i :factarr -1] > end > > do > 145 > 3.54333333333333 Thanks George for that. About the P-266: It's a job for experts to build a complete computer around that mother board. This mail is composed on my new machine though, but it will take some time before I can manage to rehabilitate all the software and made them workable. I'm sure that by then you will have completed the 3-D documentation and the implementation of SETHXYZ! Regards... [[Yehuda]] --------------------------------------------------------------- Please post messages to the Logo forum to logo-l@gsn.org. Mail questions about the list administration to logofdn@gsn.org. To unsubscribe send unsubscribe logo-l to majordomo@gsn.org.
Global SchoolNet Foundation -
Linking Kids Around the World!
Copyright GSN - All Rights Reserved
- Comments
& Questions
Visit GSN's
Global
Schoolhouse for more exciting learning resources!
Search our Site
-
Home